Gearbox Status Recognition based on TMD-SVD and POS-BP Networks Under Strong Interference

Lei He; Suqi Liu

doi:10.16578/j.issn.1004.2539.2021.05.025

您当前的位置：

首页 >

文章列表页 >

Gearbox Status Recognition based on TMD-SVD and POS-BP Networks Under Strong Interference

Development·Application | 更新时间：2022-11-10

- Gearbox Status Recognition based on TMD-SVD and POS-BP Networks Under Strong Interference
- Journal of Mechanical Transmission Vol. 45, Issue 5, Pages: 169-176(2021)
- 作者机构：
  
  1.柳州铁道职业技术学院动力技术学院，广西柳州 545000
  2.中南大学机电工程学院，湖南长沙 410083
- 作者简介：
- 基金信息：
- DOI：10.16578/j.issn.1004.2539.2021.05.025
  CLC：
扫描看全文
何雷,刘溯奇.强干扰下基于TMD-SVD和POS-BP网络的变速箱状态识别[J].机械传动,2021,45(05):169-176.

He Lei,Liu Suqi.Gearbox Status Recognition based on TMD-SVD and POS-BP Networks Under Strong Interference[J].Journal of Mechanical Transmission,2021,45(05):169-176.
何雷,刘溯奇.强干扰下基于TMD-SVD和POS-BP网络的变速箱状态识别[J].机械传动,2021,45(05):169-176. DOI： 10.16578/j.issn.1004.2539.2021.05.025.

He Lei,Liu Suqi.Gearbox Status Recognition based on TMD-SVD and POS-BP Networks Under Strong Interference[J].Journal of Mechanical Transmission,2021,45(05):169-176. DOI： 10.16578/j.issn.1004.2539.2021.05.025.

摘要

针对车辆变速箱工作环境恶劣、故障模式难以识别的问题，在现有方法基础上，提出了一种基于经验模态-小波包结合的二次模态分解（Two-layer-mode decomposition，TMD）和奇异值分解（Singular value decomposition，SVD）特征值提取方法，并结合粒子群（POS）-BP神经网络应用于变速箱故障诊断中。首先，在自行搭建的实验台上采集变速箱正常、滚动体故障、外圈裂纹、齿轮磨损4种典型状态下的振动信号；然后，用EMD分解提取信号前5个IMF分量，由于IMF,1,频谱依然较复杂，采用小波包继续进行2层分解；最终，由二次模态分解得到8个子序列，构建信号分量矩阵，再提取分量矩阵的奇异值作为特征值，将特征值输入构建好的POS-BP神经网络诊断模型中，根据输出识别变速箱故障类型。分析结果表明，该方法能有效应用于特种车辆变速箱故障诊断，诊断正确率达到92%，为复杂工况下变速箱状态识别提供了一种有效的参考途径。

Abstract

The vehicle gearbox has a bad working environment and the fault mode is difficult to identify. On the basis of existing methods， a method based on two-layer-mode decomposition （TMD） and singular value decomposition （SVD） is proposed， combined with particle swarm （POS）-BP neural network for fault diagnosis. Firstly， the vibration signals under four typical conditions of normal transmission， rolling failure， outer ring crack and gear wear are collected on a self-built experimental platform. Then， the first 5 IMFs components of the signal is decomposed by EMD， since the spectrum of IMF1 is still complicated， the wavelet packet is used to continue the 2-layer decomposition. Finally， the eight sub-sequences are obtained by TMD， and the signal component matrix is constructed. Then， the singular value （SVD） of the component matrix is extracted as the eigenvalue， the eigenvalues are entered into the constructed POS-BP neural network diagnostic model， and the gearbox fault type is identified based on the output. The analysis results show that the method can be effectively applied to the fault diagnosis of special vehicle gearboxes， and the diagnostic accuracy rate reaches 92%， which provides an effective reference for gearbox state recognition under complex conditions.

关键词

二次模态分解（TMD）奇异值分解（SVD）POS-BP神经网络故障诊断

Keywords

Two-layer-mode decomposition（TMD）Singular value decomposition（SVD）POS-BP neural networkFault diagnosis

references

贾利民，魏秀琨，秦勇，等.轨道交通列车轴承与悬挂系统故障诊断［M］.北京：人民交通出版社股份有限公司，2015：1-7.

JIA Limin，WEI Xiukun，QIN Yong，et al.Fault diagnosis for railway rolling stock bearing and vehicle suspension system［M］.Beijing：China Communications Press Co.，Ltd.，2015：1-7.

尹芳莉，谭建平，何雷，等.强冲击下变速箱滚动轴承故障诊断［J］.广西大学学报（自然科学版），2014，39（3）：620-624.

YIN Fangli，TAN Jianping，HE Lei，et al.Fault diagnosis of rolling bearing in gearbox under strong shock［J］.Journal of Guangxi University（Natural Scien-ce Edition），2014，39（3）：620-624.

李学军，何能胜，何宽芳，等.基于小波包近似熵和SVM的圆柱滚子轴承诊断［J］.振动、测试与诊断，2015，35（6）：1031-1036.

LI Xuejun，HE Nengsheng，HE Kuanfang，et al.Cylindrical roller bearing diagnosis based on wavelet packet approximate entropy and support vector machines［J］.Journal of Vibration，Testing and Diagnosis，2015，35（6）：1031-1036.

HUANG N E，SHEN Z，LONG S R，et al.The empirical mode decomposition and the Hilbert spectrum fornonlinear and non-stationary time series analysis［J］.Proceedings of the Royal Society A：Mathematical Physical and Engineering Sciences，1998，454（1971）：903-995.

张琛，赵荣珍，邓林峰，等.基于SVD-EEMD和TEO*的滚动轴承弱故障特征提取［J］.振动.测试与诊断，2019，39（4）：720-726.

ZHANG Chen，ZHAO Rongzhen，DENG Linfeng，et al.Weak fault feature extraction method for rolling based on SVD-EEMD and TEO energy spectrum［J］.Journal of Vibration，Testing and Diagnosis，2019，39（4）：720-726.

YAO G，ZHAO J J，YAO Y T，et al.Separation of systematic error based on improved EMD method［J］.Journal of Vibration and Shock，2011，33（14）：176-180.

何雷.基于局部均值分解和证据理论的变速箱故障诊断研究［D］.长沙：中南大学，2014：43-49.

HE Lei.Research on gearbox fault diagnosis based on local mean decomposition and evidence theory［D］.Changsha：Central South University，2014：43-49.

何雷，刘溯奇，蒋婷，等.基于改进LMD与BP神经网络的变速箱故障诊断［J］.机械传动，2020，44（1）：171-176.

HE Lei，LIU Suqi，JIANG Ting，et al.Gearbox fault diagnosis based on improved LMD and BP neural network［J］.Journal of Mechanical Transmission.，2020，44（1）：171-176.

陈东宁，张运东，姚成玉，等.基于FVMD多尺度排列熵和GK模糊聚类的故障诊断［J］.机械工程学报，2018，54（14）：16-27.

CHEN Dongning，ZHANG Yundong，YAO Chengyu，et al.Fault diagnosisbased on FVMD multi-scale arrangement entropy and GK fuzzy clustering［J］.Journal of Mechanical Engineering，2018，54（14）：16-27.

刘敏，张英堂，李志宁，等.基于IVMD与改进KELM的发动机故障诊断［J］.振动.测试与诊断，2019，39（4）：875-883.

LIU Min，ZHANG Yingtang，LI Zhining，et al.Engine fault diagnosis based on IVMD and improved KELM［J］.Journal of Vibration，Testing and Diagnosis.2019，39（4）：875-883.

周润景.模式识别与人工智能（基于MATLAB）［M］.北京：清华大学出版社，2018：1-10.

ZHOU Runjing.Pattern recognition and artificial intelligence （based on MATLAB）［M］.Beijing：Tsinghua University Press，2018：1-10.

卢绪祥，苏一鸣，吴家腾，等.基于EMD及灰色关联度的滑动轴承润滑状态故障诊断研究［J］.动力工程学报，2016，36（1）：42-47.

LU Xuxiang，SU Yiming，WU Jiateng，et al.Fault diagnosis on lubrication state of journal bearings based on EMD and grey relation-al degree［J］.Chinese Journal of Power Engineering，2016，36（1）：42-47.

张玲玲，肖静.基于MATLAB的机械故障诊断技术案例教程［M］.北京：高等教育出版社，2016：161-175.

ZHANG Lingling，XIAO Jing.Case study of mechanical fault diagnosis technology based on MATLAB［M］.Beijing：Higher Education Press，2016：161-175.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Bearing Feature Extraction Method Based on the Time Subsequence

A Fault Feature Extraction Method of Rolling Bearings Based on Optimized VMD and UMAP

Research on Fault Diagnosis Method of Planetary Gearboxes Based on DPD-1DCNN

Review on Fault Diagnosis Technology for High-speed EMU Gearboxes

Research on Gear Early Wear Fault Diagnosis Based on the Sparse Attention Mechanism

Related Author

No data

Related Institution

School of Information Management, Wuxi Institute of Communications Technology

IOT Engineering College, Wuxi City College of Vocational Technology

State Key Laboratory of Traction Power, Southwest Jiaotong University

School of Rail Transportation, Wuyi University

College of Mechanical and Vehicle Engineering, Taiyuan University of Technology

⁰